ORIGINAL ARTICLE
Assessment of serum-free cortisol levels in patients with adrenocortical carcinoma treated with mitotane: a pilot study
Krystallenia I. Alexandraki ** t, Gregory A. Kaltsas* t, Carel W. le Roux+, Martin Fassnacht§, Sharon Ajodha™, Mirjam Christ-Crain ** , Scott A. Akker*, William M. Drake*, Ray Edwards ], Bruno Allolio§ and Ashley B. Grossman*
*Department of Endocrinology, St Bartholomew’s Hospital, London, UK, Endocrine Division, Department of Pathophysiology, Laiko General Hospital, School of Medicine, National & Kapodistrian University of Athens, Athens, Greece, Department of Chemical Pathology, King’s College Hospital, London, UK, §Endocrinology and Diabetes Unit, Department of Medicine I, University Hospital of Wuerzburg, Wuerzburg, Germany, INETRIA, St. Bartholomew’s Hospital, London, UK and ** Department of Endocrinology, Diabetes and Clinical Nutrition, University Hospital Basel, Basel, Switzerland
Summary
Objective Mitotane treatment in adrenocortical carcinoma (ACC) results in unreliable measurement of serum total cortisol (TC) levels because of an elevation in corticosteroid-binding glob- ulin (CBG).
Design The use of a newly-developed serum-free cortisol (FC) assay was assessed to investigate the characteristics of a more valid measure of cortisol status.
Patients Sixty-two serum samples from patients with ACC trea- ted with mitotane were studied. Different subgroups were studied according to mitotane levels (<14, 14-20 and >20 mg/dl), hydro- cortisone replacement treatment, presence of Cushing’s syndrome (CS) and adrenocorticotrophin (ACTH) levels.
Measurements Serum FC was measured using a newly- developed assay, TC, CBG and plasma ACTH using conventional laboratory kits; TC-to-CBG (Free cortisol index, FCI, nmol/mg) and TC-to-FC (TFR) ratios were calculated.
Results CBG levels were elevated and positively correlated to mitotane levels. FC was positively related to TC and FCI in nearly all subgroups studied. Plasma ACTH was negatively related to parameters of cortisol levels in the total samples studied. In the ‘tar- get’ subgroup with normal ACTH levels and mitotane levels 14-20 mg/dl, no correlation of plasma ACTH with any parameter studied was seen, and FC suggested over-replacement with hydrocortisone treatment in the subgroup with CS.
Conclusions FC measurement may offer additional informa- tion in the follow-up of patients on mitotane, especially when there is a history of CS which invalidates the use of acute changes in plasma ACTH as a parameter of hydrocortisone replacement. These preliminary data suggest that it may prove useful as a bio- chemical marker when TC or FCI are invalidated by mitotane treat-
ment or plasma ACTH is suppressed by hypercortisolaemia. Larger studies are needed to substantiate the clinical utility of FC measure- ment in specific groups of patients.
(Received 4 February 2009; returned for revision 21 March 2009; finally revised 13 April 2009; accepted 11 May 2009)
Introduction
Adrenocortical carcinoma (ACC) is a rare malignancy with hetero- geneous modes of presentation and a variable but generally poor prognosis.1,2 The only adrenal-specific agent available for the treat- ment of ACC is mitotane (o’,p’-DDD), which has recently been shown to extend disease-free survival,3 although this has been ques- tioned.4-8 Mitotane is a steroidogenesis inhibitor9 which at high doses generates metabolites which bind macromolecules in adreno- cortical cell mitochondria, leading to their destruction and cellular necrosis.1º A threshold of mitotane concentration of 14 mg/l has been suggested to be necessary for an anti-tumoral response.11-14 The drug slowly accumulates, and reductions in dosage may be required to avoid toxicity, often seen with mitotane levels in excess of 20 mg/l.10,11 Therapeutic drug level monitoring is routinely per- formed in only few centres, but may be offered as service by the pharmaceutical company provider. 10-15
As mitotane affects all adrenocortical zones, glucocorticoid and mineralocorticoid replacement therapy have to be considered. Measurement of plasma adrenocorticotrophin (ACTH) or urinary- free cortisol (UFC) has been suggested to evaluate adrenal steroid secretion and adjust the dose of cortisol replacement to assure ade- quate adrenal replacement and avoid adrenal insufficiency in this situation.11 However, it is difficult to decide on a precise dose as to when such treatment should be initiated or the dose that should be adjusted as serum total cortisol (TC) measurement is unreliable because of a drug-induced increase in cortisol-binding globulin (CBG).16 Some clinicians administer hydrocortisone from the beginning of mitotane treatment; others may wait a short while for
Correspondence: Prof. Ashley B. Grossman, Department of Endocrinology, St. Bartholomew’s Hospital, London EC1A 7BE UK. Tel: +44 207 6018343; Fax: +44 207 6018505; E-mail: A.B.Grossman@qmul.ac.uk
mitotane levels to accumulate. On the other hand, the theoretical risk of mineralocorticoid deficiency does not appear to be a major clinical problem, possibly because we err in favour of hydrocorti- sone over-replacement.
Cortisol circulates in blood largely bound to CBG and albu- min,17 with the much smaller amount of unbound hormone responsible for its metabolic effects.15,18,19 It is assumed that the biologically active level of cortisol to which tissues are exposed is free cortisol (FC). Routinely available assays of adrenal function measure serum TC, but not the biologically active FC; it is usually considered that TC broadly correlates with the biologically-free fraction.
We have recently developed a serum FC assay which measures the biologically-active unbound cortisol level,20 and which is robust and simple to use. The purpose of the current study was to perform a preliminary assessment of the use of the newly- developed FC assay in patients with ACC being treated with mito- tane to see if it can provide a more valid measure of cortisol status compared to TC, and of its relationship to other markers of cortisol status such as plasma ACTH or other indices.
Subjects and methods
Patients with ACC treated with mitotane were recruited from the Department of Endocrinology, St Bartholomew’s Hospital, Lon- don, UK, and from the Endocrinology and Diabetes Unit, Depart- ment of Medicine I, University Hospital of Wuerzburg, Wuerzburg, Germany. The study was approved by the local Ethics
Committee for human studies (reference number 07/H0702/25), and written informed consent was obtained from all patients.
Blood samples were taken from the patients on arrival in the out-patient department in order to assess the levels of mitotane. The patients were advised to take their normal hydrocortisone treatment on the day of blood sampling. Blood sampling was done usually in the morning between 9:00 and 11:00 hours, 2-3 h from their morning dose of the hydrocortisone replacement. All the sam- ples collected were from patients with mitotane treatment along with hydrocortisone replacement treatment, except from the cases with persisting endogenous Cushing’s syndrome (CS) not ade- quately controlled by mitotane. The following data for each patient recruited were recorded: date of the sample, age and gender of the patient, mitotane dosage, glucocorticoid replacement treatment (type and dosage) and mitotane levels (as measured on the day of the phlebotomy). Several samples of each individual patient were accepted for the study as measurement of the mitotane level is usu- ally performed every month. A total of 3 ml of serum was stored in order to measure free and TC and CBG, and 72 samples were col- lected: 10 samples were excluded from the analysis in order to elim- inate further confounding factors as the corresponding patients were on replacement treatment with dexamethasone. Sixty-two samples from patients with age of 47.77 ± 1.51 years (36 females) were analysed. Fifty-two samples were from patients with hydro- cortisone replacement treatment and 10 samples were from patients without hydrocortisone treatment (endogenous CS). No patient received supplementary mineralocorticoid treatment. Patients under mitotane treatment were divided in three
Samples from patients treated with mitotane n = 72
Samples from patients treated with hydrocortisone replacement n = 52
Samples from patients treated with dexamethasone replacement n = 10 Excluded
Samples from patients with endogenous hydrocortisone replacement n = 10
Samples from patients with CS/without CS/NK
All samples from patients with CS
Samples from patients with mitotane levels < 14 mg/l n = 22
9
8
0
18
5
6 (5 4: ACTH within normal range)
7
5
Samples from patients with mitotane levels 14-20 mg/l n = 20
18
Samples from patients with mitotane levels > 20 mg/l n = 10
2 (ACTH within normal range)
2
5
2
9
4
1
3
4
1
8
4
Samples from patients with mitotane levels not known n = 10
2
Total studied samples
Samples with mitotane < 14 mg/l
400-000
1200-000
1200.000
Free cortisol (nmol/l)
r- = 0.87, P < 0-001
Free cortisol (nmol/l)
Free cortisol (nmol/l)
Samples from patients with normal ACTH and mitotane levels
300.000
800-000
800.000
200-000
400-000
400-000
100.000
r- = 0.85, P < 0-001
r- = 0.98, P < 0.001
0-000
0.000
0-000
0.000
2000-000
4000.000
6000.000
0.000
2000.000
4000.000
6000-000
0-000
2000.000
4000.000
6000-000
Total cortisol (nmol/l)
Total cortisol (nmol/l)
Total cortisol (nmol/l)
300-000
Free cortisol (nmol/l)
Samples from control group
Samples with mitotane = 14-20 mg/l
Subgroup with Cushing’s syndrome and without hydrocortisone treatment
600.000
1200.000
Free cortisol (nmol/l)
Free cortisol (nmol/l)
200.000
800.000
400.000
100.000
400.000
200.000
r- = 0.78, P < 0.001
r- = 0.76, P < 0-001
0.000
r- = 0.79, P = 0.006
0-000
0-000
0-000
2000-000
4000.000
6000.000
0-000
2000-000
4000.000
6000-000
0-000
2000-000
4000.000
6000.000
Total cortisol (nmol/l)
Total cortisol (nmol/l)
Total cortisol (nmol/l)
Samples from patients with Cushing’s syndrome
1200-000
Free cortisol (nmol/l)
Samples with mitotane >20 mg/l
1200.000-
Free cortisol (nmol/l)
r- = 0.93, P < 0-001
800-000
800.000-
400-000
400-000-
r- = 0.92, P < 0-001
0-000
0-000
0-000
2000-000
4000.000
6000-000
0-000
2000.000
4000.000
6000-000
Total cortisol (nmol/l)
Total cortisol (nmol/l)
subgroups: 22 samples, from patients with sub-therapeutic levels of mitotane (less than 14 mg/l); 20 samples, from patients with levels of mitotane in the therapeutic range (14-20 mg/l); 10 samples, from patients over-treated (>20 mg/l).11 Plasma ACTH levels were determined in 11, 13 and 6 samples of these groups, respectively. The subgroup of patients with normal ACTH levels (defined as 10- 60 ng/l) and mitotane levels within the therapeutic range was anal- ysed separately, as it was considered to represent the ‘target’ group which satisfied the mitotane therapeutic target without evidence of adrenocortical insufficiency. The samples were also divided into those from patients with CS and in those from patients without CS (Fig. 1). A small group of 18 healthy subjects [age: 46-33 ± 3.40 years (P = 0-74 vs. patients); 10 females (P = 0.85 compared to patients)], which had been recruited for another published study,20 was used in order to determine the normal reference range compared to the patient population.
Serum FC was measured in-house, by a recently-developed assay, as previously described.20 The fractions of free and bound cortisol in serum samples were first separated by equilibrium dialy- sis, followed by measurement of the concentration of the free frac- tion in the dialysate using an enzyme immunoassay (EIA, NETRIA®, London, UK). Dialysate containing the free fraction
was assayed by EIA. The calculated sensitivity of the EIA was 0-4 nmol/l (at 2-5 SDs from zero cortisol concentration). The inter- assay coefficients of variance (CV) were 3-5% at 2.1 nmol/1, 6.8% at 24 nmol/l and 10-0% at 52 nmol/l. The intra-assay CV was assessed by precision profile with the following CV: 10% at 1 nmol/l, 6% at 10 nmol/l, 5% at 20 nmol/l and 7% at 100 nmol/l. TC levels were measured using an enzyme immunoassay (EIA) kit (NETRIA; St. Bartholomew’s Hospital, London). The inter-assay CV at 49, 411 and 729 nmol/l was 10%, 5-3% and 9-1%, respectively. The intra- assay CV at 50, 250 and 750 nmol/l was 6.9%, 4.3% and 6-5%, respectively. The calculated sensitivity of the assay was 25 nmol/l. Plasma ACTH (ng/l) was assayed using the Immulite 2500 method, which is a fully automated two-site non-extraction chemilumines- cent immunometric assay. The %CV is <7% at both 30 and 430 ng/1, with an assay range of 5-1250 ng/l and an analytical sen- sitivity of 5 ng/l. For CBG measurement, all serum samples were assayed in duplicate using a DRG Diagnostics radioimmunoassay kit. The assay had an inter- and intra-assay percentage (%) below 10% and a detection limit of 1·3 mg/l. Mitotane levels were deter- mined using an HPLC method [Lysodren (HRA Pharma, Paris, France); PAREXEL] in 52 samples (10 samples were provided from patients under mitotane treatment, but the mitotane levels were
Total studied samples
Free cortisol (nmol/l)
Free cortisol (nmol/l)
Samples with mitotane < 14 mg/l
Free cortisol (nmol/l)
Subgroup with Cushing’s syndrome and without hydrocortisone
1200-000
r- = 0.78, P < 0.001
1200.000
1200.000
treatment-
800-000
800-000
800.000
400-000
400-000
400.000
r- = 0.80, P < 0.001
r- = 0.70, P < 0.03
0-000
0-000
0.000
0.000
25-000
50.000
75.000
100-000
0-000
25-000
50-000
75.000
100.000
0-000
25.000
50.000
75.000
100.000
Free cortisol index (nmol/mg)
Free cortisol index (nmol/mg)
Free cortisol index (nmol/mg)
750.000
Free cortisol (nmol/l)
Samples from patients with normal ACTH and mitotane levels
1500.000-
Free cortisol (nmol/l)
Samples with mitotane = 14-20 mg/l
300.000
Free cortisol (nmol/l)
Samples from patients with Cushing’s syndrome
1000-000-
200.000
500-000
100.000
500.000
250-000
r- = 0.58, P < 0-008
r- = 0.97, P < 0.001
r- = 0.90, P < 0-001
0.000
0.000
25-000
50-000
75.000
100.000
0.000
0-000
Free cortisol index (nmol/mg)
0-000
25.000
50.000
75.000
100.000
0-000
25-000
50-000
75-000
100.000
Free cortisol index (nmol/mg)
Free cortisol index (nmol/mg)
400.000
Samples with mitotane > 20 mg/l
1500.000
Free cortisol (nmol/l)
Free cortisol (nmol/l)
Samples from patients without Cushing’s syndrome
300.000
1000-000
r- = 0.58, P < 0.004
200.000
500-000
100.000
r- = 0.97, P < 0.001
0-000
0-000
0-000
25-000
50-000
75.000
100.000
Free cortisol index (nmol/mg)
0.000
25-000
50.000
75.000
Free cortisol index (nmol/mg)
100.000
not determined or were lost). The inter- and intra-assay CVs of the method are within 20% at the limit of quantification level and within 15% at the higher levels. The lower limit of quantification was 1-01 g/ml, with an assay range from 1-01 to 49-5 µg/ml.
The Free Cortisol index (FCI, nmol/mg) was calculated using the following formula: TC (nmol/l)/CBG (mg/l).21 The TC-to-FC ratio (TFR) was also calculated.
Statistical analysis
The results are reported as mean values ± SE. Statistical significance in the results were accepted at a P-value < 0-05. The normal distri- bution of continuous variables was assessed by applying the non- parametric Kolmogorov-Smirnov test. Independent samples t-test was used for two groups’ comparison and the Mann-Whitney U-test for variables which were not normally distributed. Multiple analysis of variance (ANOVA) was used for multiple group compari- son analysis (with Bonferroni correction) and the Kruskal-Wallis ANOVA for variables which were not normally distributed. Correla- tions between variables were evaluated by Pearson’s coefficient except for variables not normally distributed, or with less than 10 values, which were evaluated by Spearman’s correlation coefficient.
Results
Grouping the 62 samples together, CBG was elevated above the normal range and was positively correlated with mitotane levels (r = 0-51, P < 0-001). FC was also positively related to TC and FCI in the total group as well as in the subgroups (Figs 2 and 3).
In subgroups mitotane <14, 14-20 and 20 mg/l, serum FC levels were 148.87 ± 66-03, 68.88 ± 13.88, 166.15 ± 83.36 nmol/l, respectively, above the range found in the normal population (36.94 ± 4.31 nmol/l, range: 16-34-67-55), but without showing statistically significant differences from controls or between groups. Serum TC levels were: 1010.11 ± 310-35, 949-61 ± 282.66 and 1102.84 ± 432.79 nmol/l, respectively, all elevated and not different from each other. FCI did not differ between the subgroups, ACTH levels were suppressed in the group with mitotane <14 mg/dl com- pared to group with levels 14-20 mg/dl (P = 0-02) and the con- trols. CBG levels were elevated in all patients receiving mitotane compared to the controls, correlating with the mitotane levels (Table 1).
Eleven patients had plasma ACTH (2.73 ± 0-71 pmol/l, range: 1.11-8.22) and mitotane levels (16.6 ± 0-52 mg/l, range: 14-20) in the target range, and they also had FC (62.10 ± 19.91, P = 0.24)
| Variables | Total samples studied (N = 62) | Samples with mitotane < 14 mg/l (N = 22) | Samples with mitotane = 14-20 mg/l (N = 20) | Samples with mitotane > 20 mg/l (N = 10) | Samples from control group (N = 18) |
|---|---|---|---|---|---|
| Free cortisol (nmol/l) | 137-09 ± 30-03* | 148-87 ± 66.03 | 68.88 ± 13.88 | 166.15 ± 83.36 | 36.94 ± 4·31 |
| Total cortisol (nmol/l) | 1031-59 ± 172.55 | 1010-11 ± 310-35 | 949-61 ± 282-66 | 1102-84 ± 432.79 | 488-94 ± 32.23 |
| TFR | 10-79 ± 0-85* | 11-47 ± 1.24 | 13.80 ± 1.84 *** | 8:21 ± 0.97 | 14.77 ± 1.09 |
| Mitotane levels (mg/l) | 14.96 ± 1.02 | 8.36 ± 0.94 | 16.55 ± 0-42 | 25.66 ± 1.13 | (-) |
| ACTH (pmol/l) | 13.71 ± 7.65* | 1.92 ± 0-48*, ** | 26.12 ± 19-56 | 14.18 ± 6.84 | 6.01 ± 1:00 |
| FCI (nmol/mg) | 8.44 ± 2.06* | 10.15 ± 4.66*, | 6.91 ± 2.63 | 8.64 ± 5.87 | 13.86 ± 1.65 |
| CBG (µg/ml) | 169-28 ± 9.79* | 137:35 ± 14.24*, *** | 172.595 ± 11.78*, *** | 251.45 ± 10.92* | 37.38 ± 4.00 |
TFR, total-to-free cortisol ratio; ACTH, adrenocorticotrophin; FCI, free cortisol index; CBG, cortisol-binding protein.
Values ± SE; * P ≤ 0.05 vs. controls; ** P ≤ 0-05 vs. group with mitotane 14-20 mg/dl; *** P ≤ 0.05 vs. group with mitotane >20 mg/dl.
| Variables | Samples from patients with normal ACTH and mitotane levels (N = 11) | Samples from patients with normal ACTH, mitotane levels and Cushing's syndrome (N = 8) | Samples from patients with normal ACTH, mitotane levels without Cushing's syndrome (N = 3) | Samples from patients with hydrocortisone treatment (N = 52) | Samples from patients with Cushing's syndrome and without hydrocortisone treatment (N = 10) | Samples from control group (N = 18) |
|---|---|---|---|---|---|---|
| Free cortisol (nmol/l) | 62-10 ± 19.91 | 62-41 ± 29-87 | 43.98 ± 20-08 | 73.71 ± 10-38* | 466-63 ± 142.56* | 36.94 ± 4.31 |
| Total cortisol (nmol/l) | 891-92 ± 347-31 | 879.59 ± 510-37 | 508.60 ± 244-58 | 695-70 ± 115.91 | 2744.66 ± 663-43* | 488-94 ± 32-23 |
| TFR | 13.65 ± 1-42 | 12.59 ± 1.52 | 14.47 ± 3.89 | 11.35 ± 0-93 | 9.00 ± 1.92* | 14.77 ± 1.09 |
| Mitotane (mg/l) | 16.56 ± 0.52 | 16.89 ± 0.73 | 15-47 ± 0-61 | 15.67 ± 1.05 | 10-54 ± 3.15 | (-) |
| ACTH (pmol/l) | 2.73 ± 0.71* | 1.83 ± 1.01* | 1.13 ± 3.61 | 78.67 ± 45.17* | 2.04 ± 0·63* | 6.01 ± 1.00 |
| FCI (nmol/mg) | 5-98 ± 2.42* | 6.04 ± 3.62* | 3.26 ± 1.30* | 4.83 ± 1.05* | 27.68 ± 9-89 | 13.86 ± 1.65 |
| CBG (µg/ml) | 165.25 ± 10-65* | 177-11 ± 13.36* | 140-90 ± 18.79* | 177.80 ± 10-79* | 124-98 ± 18.31* | 37:38 ± 4.00 |
ACTH, adrenocorticotrophin; TFR, total-to-free cortisol ratio; FCI, free cortisol index; CBG, cortisol-binding protein. Values ± SE; * P ≤ 0-05 vs. controls.
| Variables | Samples from patients with Cushing's syndrome (N = 31) | Samples from patients without Cushing's syndrome (N = 23) | Samples from control group (N = 18) |
|---|---|---|---|
| Free cortisol (nmol/l) | 192.11 ± 57.08* | 83.293 ± 18.61 | 36.94 ± 4.31 |
| Total cortisol (nmol/l) | 1364.85 ± 302.16* | 678-22 ± 199-48 | 488.94 ± 32-23 |
| TFR | 10.78 ± 0-89 | 10.91 ± 1.79 | 14.77 ± 1.09 |
| Mitotane levels (mg/l) | 13.07 ± 1.34 | 15.89 ± 1.90 | (-) |
| ACTH (pmol/l) | 1.84 ± 0.29*, ** | 37.62 ± 25.01+ | 6.01 ± 1.00 |
| FCI (nmol/mg) | 11.93 ± 3.85* | 5.62 ± 2.12* | 13.86 ± 1.65 |
| CBG (µg/ml) | 159.85 ± 11.62* | 163.76 ± 14.57* | 37.38 ± 4.00 |
TFR, total-to-free cortisol ratio; ACTH, adrenocorticotrophin; FCI, free cortisol index; CBG, cortisol-binding protein.
Values ± SE; * P ≤ 0-05 vs. controls; ** P ≤ 0.05 vs. patients without CS;+P < 0.1 vs. controls.
and TC (891.92 + 347.31, P = 0.27) not statistically different from controls, but the FCI (P = 0-02) was statistically lower. CBG levels were significantly elevated (P < 0-001) compared to controls. In this target group, eight patients presented with CS and three patients without CS; no difference was observed between those two
groups in any parameter studied (Table 2). Ten patients were not receiving glucocorticoid treatment at the time of the study: the parameters studied are as shown in Table 2.
Regarding the comparison between the groups depending on the presence or absence of CS, these did not differ in FC (P = 0-57)
| Correlation coefficient (r) | P-value | |
|---|---|---|
| Total studied samples | ||
| FC vs. | ||
| ACTH | -0-45 | 0-008 |
| ACTH vs. | ||
| Total cortisol | -0-49 | 0-003 |
| FCI | -0-43 | 0-01 |
| Mitotane levels | +0-37 | 0-047 |
| CBG | +0-51 | <0-001 |
| CBG vs. | ||
| Mitotane levels | +0-51 | <0-001 |
| Mitotane subgroups | ||
| Samples with mitotane 14-20 mg/l | ||
| ACTH vs. | ||
| Total cortisol | -0-55 | 0-05 |
| FCI | -0-56 | 0-05 |
| TFR vs. | ||
| FCI | +0-70 | 0-001 |
| CBG | +0-61 | 0-004 |
| Samples from patients with Cushing's syndrome | ||
| FC vs. | ||
| ACTH | -0-44 | 0-055 |
| Samples from patients without Cushing's syndrome | ||
| ACTH vs. | 0-02 | |
| FCI | -0-71 | |
| FCI vs. | ||
| TFR | 0-69 | <0-001 |
| CBG vs. | ||
| TFR | -0-48 | 0-02 |
| Mitotane | +0-89 | <0-001 |
ACTH, adrenocorticotrophin; FCI, free cortisol index; CBG, cortisol- binding protein; TFR, total-to-free cortisol ratio. P ≤ 0.05.
levels, TC (P = 0.13), FCI (P = 0.13), TFR (P= 1.00), mitotane (P = 0.50) or CBG levels (P = 1.00); in the CS group, lower ACTH plasma levels were observed (P = 0-02) (Table 3).
The correlations of the parameters studied in each subgroup are presented in Table 4.
Discussion
This study clearly demonstrated a significant CBG rise in parallel with mitotane levels in patients treated for ACC, producing an expected rise in TC and a drop in FCI. FC levels measured by the recently developed assay showed a highly positive correlation with TC and FCI levels in the total population and the subgroups sam- pled. This finding might imply that FC assessment is redundant in these patients. However, FC assessment provides additional clinical information to ACTH in particular conditions, such as patients with endogenous hypercortisolaemia, both in the early stages of the disease as well as in cases of recurrence or before the attainment of therapeutic mitotane values when ACTH levels are still suppressed.
In terms of the monitoring of patients on mitotane, there are important considerations regarding the choice of the optimal
marker. Mitotane leads to a rise in CBG,16 invalidating the use of TC in monitoring the hypercortisolaemia in patients with ACC and it also increases the metabolic clearance of glucocorticoids15 such as dexamethasone, rendering its use hampered by uncertainty as to dosage. Plasma levels of ACTH have been considered to be useful in the assessment of glucocorticoid replacement adequacy.22 However, measurement of plasma ACTH is pulsatile and results may not be readily and rapidly available;23 furthermore, patients with ACC with long-standing hypercortisolism have suppressed levels of ACTH, but it may be unclear whether this is due to over- treatment or to persistent pituitary-adrenal suppression. On the other hand, UFC remains an alternative but requires an accurate 24 h collection by the patient, and may be problematic on a regular basis as difficulties in such collections are well reported.24 Finally, salivary FC sampling is promising,25 requiring only a small addi- tional effort in its collection, although problems may still arise.24 Therefore, the concept of an accurate marker, measured on the same sample as mitotane levels that does not require any extra effort from the patient, and which can be measured at their follow- up during mitotane monitoring, seems an attractive solution for this specific population of patients which is already stressed by a disease with low survival rates. FC may prove to be a useful addi- tional assessment when mitotane treatment makes TC and FCI less reliable, and would obviate the necessity for a 24-h urine collection.
In cases of functional tumours with CS, ACTH levels can be undetectable (remain suppressed by endogenous hypercortiso- lism), and remain low for some time following the lowering of cor- tisol levels. In such cases, neither TC nor ACTH can be used to adequately assess glucocorticoid sufficiency, suggesting the neces- sity of another marker.
However, while FC correlated with most of the markers of cortisol levels, it is difficult to be sure that is added useful clinical informa- tion. In patients with CS, ACTH levels may remain persistently sup- pressed, and our finding that FC levels were still elevated above the normal range might suggest that such patients were still exposed to excessive glucocorticoids; however, there is no ‘gold standard’ to assess tissue exposure (Table 3). In the total population, FC levels suggested that in general, patients are adequately substituted, and possibly even over-substituted (Table 2). While random samples in patients on hydrocortisone therapy may be of limited value, the FC results would indicate that such patients are not adrenally-insuffi- cient. A particularly important subgroup is the one in which mito- tane achieved its therapeutic range target, while plasma ACTH remains in the low-normal range suggesting adequate glucocorti- coid replacement. In this group, no correlation was found between FC and ACTH levels, but this may again reflect the persistence in ACTH suppression seen in long-standing patients with recovery from hypercortisolaemia. In the target subgroup without CS, the group was unfortunately too small to enable valid comparisons to be made.
This study has several limitations, such as the small sample num- ber, the absence of a definite subgroup of patients in different stages of ACC, the fact that parameters studied were not known in all the subgroups studied, and that more than half the patients had a his- tory of CS, making the differentiation of endogenous excess and exogenous replacement more difficult. While we examined patients
from two major centres with a special interest in ACC, ACC is a rare disease and taking into account variables such as the presence or absence of CS, the level of mitotane and the use or otherwise of hydrocortisone or other replacement therapy, it is difficult to obtain large sample numbers for each subgroup analysis. Neverthe- less, the aim of this project was the pilot investigation of the appli- cability of this newly-developed serum FC assay in this population.
In summary, measurement of FC appears to be a promising bio- chemical marker for assessing glucocorticoid sufficiency in those patients when TC or FCI are invalidated by mitotane use, particu- larly in cases of CS with suppressed ACTH. Larger studies with the inclusion of clinical parameters and a more complete biochemical profile may aid in the determination of the clinical utility of FC measurement compared to plasma ACTH or UFC in specific groups of patients. Furthermore, a comparative study with salivary cortisol might shed light on the best marker of cortisol levels in this patient population.
Competing interests/financial disclosure
Nothing to declare.
References
1 Wajchenberg, B., Albergaria, P.M., Medonca, B. et al. (2000) Adre- nocortical carcinoma: clinical and laboratory observations. Cancer, 88, 729-737.
2 Dackiw, A.P., Lee, J.E., Gagel, R.F. et al. (2001) Adrenal cortical carcinoma. World Journal of Surgery, 25, 914-926.
3 Terzolo, M., Angeli, A., Fassnacht, M. et al. (2007) Adjuvant mito- tane treatment for adrenocortical carcinoma. New England Journal of Medicine, 356, 2372-2380.
4 Schteingart, D.E. (2007) Adjuvant mitotane therapy of adrenal can- cer - use and controversy. New England Journal of Medicine, 356, 2415-2418.
5 Bertherat, J., Coste, J. & Bertagna, X.N. (2007) Adjuvant mitotane in adrenocortical carcinoma. New England Journal of Medicine, 357, 1256-1257. author reply 1259.
6 Dickstein, G., Shechner, C. & Nativ, O. (2007) Adjuvant mitotane in adrenocortical carcinoma. New England Journal of Medicine, 357, 1257-1258. author reply 1259.
7 Machens, A. & Dralle, H. (2007) Adjuvant mitotane in adrenocor- tical carcinoma. New England Journal of Medicine, 357, 1258-1259. author reply 1259.
8 Lee, J.E. (2007) Adjuvant mitotane in adrenocortical carcinoma. New England Journal of Medicine, 357, 1258. author reply 1259.
9 Ojima, M., Saitoh, M., Itoh, N. et al. (1985) The effects of o,p’- DDD on adrenal steroidogenesis and hepatic steroid metabolism. Nippon Naibunpi Gakkai Zasshi, 61, 168-178.
10 Allolio, B. & Fassnacht, M. (2006) Clinical review: adrenocortical carcinoma: clinical update. Journal of Clinical Endocrinology and Metabolism, 91, 2027-2037.
11 Schteingart, D.E., Doherty, G.M., Gauger, P.G. et al. (2005) Man- agement of patients with adrenal cancer: recommendations of an
international consensus conference. Endocrine-Related Cancer, 12, 667-680.
12 van Slooten, H., Moolenaar, A.J., van Seters, A.P. et al. (1984) The treatment of adrenocortical carcinoma with o,p-DDD: prognostic implications of serum level monitoring. European Journal of Cancer Clinical Oncology, 20, 47-53.
13 Haak, H.R., Hermans, J., van de Velde, C.J. et al. (1994) Optimal treatment of adrenocortical carcinoma with mitotane: results in a consecutive series of 96 patients. British Journal of Cancer, 69, 947- 951.
14 Baudin, E., Pellegriti, G., Bonnay, M. et al. (2001) Impact of moni- toring plasma 1,1-dichlorodiphenildichloroethane (o,p_DDD) levels on the treatment of patients with adrenocortical carcinoma. Cancer, 92, 1385-1392.
15 Hague, R.V., May, W. & Cullen, D.R. (1989) Hepatic microsomal enzyme induction and adrenal crisis due to o, p _- DDD therapy for metastatic adrenocortical carcinoma. Clinical Endocrinology, 31, 51-57.
16 van Seters, A.P. & Moolenaar, A.J. (1991) Mitotane increases the blood levels of hormone-binding proteins. Acta Endocrinologica, 124, 526-533.
17 Dunn, J.F., Nisula, B.C. & Rodbard, D. (1981) Transport of steroid hormones: binding of 21 endogenous steroids to both testosterone- binding globulin and corticosteroid-binding globulin in human plasma. Journal of Clinical Endocrinology and Metabolism, 53, 58-68.
18 Brien, T.G. (1981) Human corticosteroid binding globulin. Clinical Endocrinology, 14, 193-212.
19 Ogawa, K., Sueda, K. & Matsui, N. (1983) The effect of cortisol, progesterone, and transcortin on phytohemagglutinin-stimulated human blood mononuclear cells and their interplay. Journal of Clinical Endocrinology and Metabolism, 56, 121-126.
20 Christ-Crain, M., Jutla, S., Widmer, I. et al. (2007) Measurement of serum free cortisol shows discordant responsivity to stress and dynamic evaluation. Journal of Clinical Endocrinology and Metabo- lism, 92, 1729-1735.
21 le Roux, C.W., Chapman, G.A., Kong, W.M. et al. (2003) Free cortisol index is better than serum total cortisol in determining hypothalamic-pituitary-adrenal status in patients undergoing surgery. Journal of Clinical Endocrinology and Metabolism, 88, 2045-2048.
22 Feek, C.M., Ratcliffe, J.G., Seth, J. et al. (1981) Patterns of plasma cortisol and ACTH concentrations in patients with Addison’s dis- ease treated with conventional corticosteroid replacement. Clinical Endocrinology, 14, 451-458.
23 Arnaldi, G., Angeli, A., Atkinson, A.B. et al. (2003) Diagnosis and complications of Cushing’s syndrome: a consensus state- ment. Journal of Clinical Endocrinology and Metabolism, 88, 5593-5602.
24 Vilar, L., Freitas Mda, C., Faria, M. et al. (2007) Pitfalls in the diag- nosis of Cushing’s syndrome. Arquivos brasileiros de endocrinologia e metabologia, 51, 1207-1216.
25 Yaneva, M., Mosnier-Pudar, H., Dugué, M.A. et al. (2004) Mid- night salivary cortisol for the initial diagnosis of Cushing’s syn- drome of various causes. Journal of Clinical Endocrinology and Metabolism, 89, 3345-3351.